1

Modelling Hydrological Processes in Arid and Semi Arid Areas

- an Introduction to the Workshop

Howard S. Wheater

Department of Civil and Environmental Engineering

Imperial College of Science, Technology and Medicine,
London SW7 2BU, UK

INTRODUCTION

In the arid and semi-arid regions of the world, water resources are limited, and under severe
and increasing pressure due to expanding populations, increasing per capita water use and
irrigation. Point and diffuse pollution, increasing volumes of industrial and domestic waste
and over-abstraction of groundwater provide a major threat to those scarce resources. Floods
are infrequent, but extremely damaging, and the threat from floods to lives and infrastructure
is increasing due to urban development. Ecosystems are fragile, and under threat from
groundwater abstractions and the management of surface flows. Added to these pressures is
the uncertain threat of climate change. Clearly effective water management is essential, and
this requires appropriate decision support systems, including modelling tools.

Modelling methods have been widely used over 40 years for a variety of purposes, but almost
all modelling tools have been primarily developed for humid area applications. Arid and
semi-arid areas have particular challenges that have received little attention. One of the
primary aims of this workshop is to bring together world-wide experience and some of the
world’s leading experts to provide state-of-the-art guidance for modellers of arid and semi-
arid systems.

The development of models has gone hand-in-hand with developments in computing power.
While event-based models originated in the 1930s and could be used with hand calculation,
the first hydrological models for continuous simulation of rainfall-runoff processes emerged
in the 1960s, when computing power was sufficient to represent all of the land-phase
processes in a simplified, ‘conceptual’ way. Later, in the 1970s and ‘80s, increases in power
enabled ‘physically-based’ hydrological models to be developed, solving a coupled set of
partial differential equations to represent overland, in-stream and subsurface flow and
transport processes, together with evaporation from land and water surfaces. And currently,
global climate models are able to represent the global hydrological cycle with simplified
physics-based models. In parallel, recent developments in computer power provide the ability
to use increasingly powerful methods for the analysis of model performance and to specify
the uncertainty associated with hydrological simulations. There have as a result been
important developments in our understanding of modelling strengths and limitations. The
workshop will present a range of modelling approaches and introduce methods of uncertainty
analysis.

The relationship between models and data is fundamental to the modelling task. Current
technology and computing power can provide powerful pre- and post-processors for
hydrological models through Geographic Information Systems, linking with digital data sets
to provide a user-friendly modelling environment. Some of these methods will be
 2

demonstrated here, and an important issue for discussion is the extent to which such methods
are applicable to data-sparse environments, and for countries where the underlying digital
data may be hard to obtain. Global developments in remote sensing, coupled with modelling
and data assimilation, are providing new sources of information. For example, precipitation
estimates for mid-latitudes are now available in near real-time; remote sensing of water body
elevation is approaching the point where resolution is useful for real-time hydrological
modelling. Again, the workshop will illustrate new data products and discuss their
applicability (see the workshop paper by Soroosh Sorooshian et al.).

The workshop has been supported by UNESCO and co-sponsors under the G-WADI
initiative. G-WADI seeks to provide a global forum for the exchange of experience,
information and tools. We expect that the formal presentations will benefit greatly from the
wide experience of the workshop participants, and the material presented here will be refined
in the light of participants’ comments and produced on the web in due course. Above all we
hope that the material will provide insight and tools to help practitioners world-wide.

In this introductory talk I will aim to set the scene with a perspective on the strengths and
weaknesses of alternative modelling approaches, the special features of arid areas, and the
consequent modelling challenges.

RAINFALL-RUNOFF MODELLING

The workshop presupposes a basic understanding of modelling, and for those requiring more
introductory material, the text-book by Beven (2000) provides an excellent introduction, and
several recent advanced texts are also available (e.g. Wagener et al., 2004; Duan et al., 2003,
Singh and Frevert, 2002a,b.). Nevertheless a brief introduction to modelling terminology and
issues is included here, to provide a common framework for subsequent discussion.

A model is a simplified representation of a real world system, and consists of a set of

simultaneous equations or a logical set of operations contained within a computer program.
Models have parameters which are numerical measures of a property or characteristics that
are constant under specified conditions. A lumped model is one in which the parameters,
inputs and outputs are spatially averaged and take a single value for the entire catchment. A
distributed model is one in which parameters, inputs and outputs vary spatially. A semi-
distributed model may adopt a lumped representation for individual subcatchments. A model
is deterministic if a set of input values will always produce exactly the same output values,
and stochastic if, because of random components, a set of input values need not produce the
same output values. An event-based model produces output only for specific time periods,
whereas a continuous model produces continuous output.

The tasks for which rainfall-runoff models are used are diverse, and the scale of applications
ranges from small catchments, of the order of a few hectares, to that of global models.
Typical tasks for hydrological simulation models include:-

• Modelling existing catchments for which input-output data exist,

o e.g. Extension of data series for flood design of water resource evaluation,
operational flood forecasting, or water resource management
• Runoff estimation on ungauged basins
• Prediction of effects of catchment change
e.g. Land use change, climate change
 3

• Coupled hydrology and geochemistry

e.g. Nutrients, Acid rain
• Coupled hydrology and meteorology
e.g. Global Climate Models

Clearly, the modelling approach adopted will in general depend on the required scale of the
problem (space-scale and time-scale), the type of catchment, and the modelling task. Some of
the tasks pose major challenges, and it is helpful to consider a basic classification of model
types, after Wheater et al. (1993), and their strengths and weaknesses.

Metric models
At the simplest level, all that is required to reproduce the catchment-scale relationship
between storm rainfall and stream response to climatic inputs is a volumetric loss, to account
for processes such as evaporation, soil moisture storage and groundwater recharge, and a time
distribution function, to represent the various dynamic modes of catchment response. This is
the basis of the unit hydrograph method, developed in the 1930s, which, in its basic form,
represents the stream response to individual storm events by a non-linear loss function and
linear transfer function. The simplicity of the method provides a powerful tool for data
analysis. Once a set of assumptions has been adopted (separating fast and slow components
of the streamflow hydrograph and allocating rainfall losses), rainfall and streamflow data can
be readily analysed, and a unique model determined.

This analytic capability has been widely used in regional analysis. In the UK, for example,
the 1975 Flood Studies Report (NERC, 1975) used data from 138 UK catchments to define
regression relationships between the model parameters and storm and catchment
characteristics for the rainfall loss and transfer functions. This lumped, event-based model
provides the basic tool for current UK flood design, and, through the regional regression
relationships, a capability to model flow on ungauged catchments (the regional relationships
were updated in the 1999 Flood Estimation Handbook (Institute of Hydrology, 1999) through
the replacement of manual by digital map-based characteristics).

The unit hydrograph is also widely adopted internationally in the form of the US Soil
Conservation Service model, available within the US Corps of Engineers HEC1 model, as
described in the workshop paper by Radwan Al-Weshah. Synthetic unit hydrographs can
readily be generated based on default model parameters, which is particularly helpful in data-
scarce situations. However, relatively little work has been done to evaluate the associated
uncertainty with these estimates.

This data-based approach to hydrological modelling has been defined as metric modelling
(Wheater et al., 1993). The essential characteristic of metric models is that they are based
primarily on observations and seek to characterise system response from those data. In
principle, such models are limited to the range of observed data, and effects such as
catchment change cannot be directly represented. In practice, the analytical power of the
method has enabled some effects of change to be quantified; the UK regional analysis found
the degree of urban development to be an important explanatory variable, and this is used in
design to mitigate impacts of urbanisation.

The unit hydrograph is a simple, event, model with limited performance capability. However
methods of time-series analysis can be used to identify more complex model structures for
event or continuous simulation. These are typically based on parallel linear stores, and
 4

provide a capability to represent both fast and slow-flow components of a streamflow

hydrograph (see for example Barry Croke and Tony Jakeman’s workshop paper). These
provide a powerful set of tools for use, with updating techniques, in real-time flood
forecasting, as we shall hear later from Prof. Peter Young.

Conceptual models
The most common class of hydrological model in general application incorporates prior
information in the form of a conceptual representation of the processes perceived to be
important. The model form originated in the 1960s, when computing power allowed, for the
first time, integrated representation of the terrestrial phase of the hydrological cycle, albeit
using simplified relationships, to generate continuous flow sequences. These conceptual
models are characterised by parameters that usually have no direct, physically measurable
identity. The Stanford Watershed Model (Crawford and Linsley, 1966) is one of the earliest
examples, and, with some 16-24 parameters, one of the more complex. To apply these models
to a particular catchment, the model must be calibrated, i.e. fitted to an observed data set to
obtain an appropriate set of parameter values, using either a manual or automatic procedure.
Many of the models presented in the workshop (e.g. by Denis Hughes, George Leavesley et
al., KD Sharma, RD Singh) fall into this category.

The problem arises with this type of model that the information content of the available data
is limited, particularly if a single performance criterion (objective function) is used (see
Kleissen et al., 1990) and hence in calibration the problem of non-identifiability arises,
defined by Beven (1993) as "equifinality". For a given model, many combinations of
parameter values may give similar performance (for a given performance criterion), as indeed
may different model structures. This has given rise to two major limitations. If parameters
cannot be uniquely identified, then they cannot be linked to catchment characteristics, and
there is a major problem in application to ungauged catchments. Similarly, it is difficult to
represent catchment change if the physical significance of parameters is ambiguous.

Developments in computing power, linked to an improved understanding of modelling

limitations, have led to some important theoretical and practical developments for conceptual
modelling. Firstly, recognising the problem of parameter ambiguity, appropriate methods to
analyse and represent this have been developed. The concept of Generalized Sensitivity
Analysis was introduced (Spear and Hornberger, 1980), in which the search for a unique best
fit parameter set for a given data set is abandoned; parameter sets are classified as either
"behavioural" (consistent with the observed data) or "non-behavioural" according to a defined
performance criterion. An extension of this is the Generalised Likelihood Uncertainty
Estimation (GLUE) procedure (Beven and Binley, 1992; Freer et al., 1996). Using Monte
Carlo simulation, parameter values are sampled from the feasible parameter space
(conditioned on prior information, as available). Based on a performance criterion, a
"likelihood" measure can be evaluated for each simulation. Non-behavioural simulations can
be rejected (based on a pre-selected threshold value), and the remainder assigned re-scaled
likelihood values. The outputs from the runs can then be weighted and ranked to form a
cumulative distribution of output time series, which can be used to represent the modelling
uncertainty. This formal representation of uncertainty is an important development in
hydrological modelling practice, although it should be noted that the GLUE procedure lumps
together various forms of uncertainty, including data error, model structural uncertainty and
parameter uncertainty. More generally, Monte Carlo analysis provides a powerful set of
methods for evaluating model structure, parameter identifiability and uncertainty. For
example, in a recent refinement (Wagener et al., 2003a,b), parameter identifiability is
 5

evaluated using a moving window to step through the output time-series, thus giving insight
into the variability of model performance with time.

A second development is a recognition that much more information is available within an

observed flow time series than is indicated by a single performance criterion, and that
different segments of the data contain information of particular relevance to different modes
of model performance (Wheater et al., 1986). This has long been recognised in manual
model calibration, but has only recently been used in automatic methods. A formal
methodology for multi-criterion optimisation has been developed for rainfall-runoff
modelling (e.g. Gupta et al., 1998, Wagener et al., 2000, 2002). Provision of this additional
information reduces the problem of equifinality (although the extent to which this can be
achieved is an open research issue), and provides new insights into model performance. For
example, if one parameter set is appropriate to maximise peak flow performance, and a
different set to maximise low flow performance, this may indicate model structural error, or
in particular that different models apply in different ranges. Modelling tool-kits for model
building and Monte-Carlo analysis are currently available, which include GLUE and other
associated tools for analysis of model structure, parameter identifiability, and prediction
uncertainty (Lees and Wagener, 1999; Wagener et al., 1999).

An important reason for detailed analysis of model structure and parameter identifiability is
to explore the trade-off between identifiability and performance to produce an optimum
model (or set of models) for a particular application. Thus for regionalisation, the focus
would be on maximising identifiability (i.e. minimising parameter uncertainty), so that
parameters can be related to catchment characteristics.

In several senses, therefore, current approaches to parsimoneous conceptual modelling

represent an extension of the metric concept (and have thus been termed hybrid metric-
conceptual models). There has been a progressive recognition that the 1960s first-generation
conceptual models, while seeking a comprehensive and integrated representation of the
component processes, are non-identifiable. The current generation of stochastic analysis tools
allows detailed investigation of model structure and parameter uncertainty, leading to
parameter-efficient models that seek to extract the maximum information from the available
data. They also allow formal recognition of uncertainty in model parameters, and provide the
capability to produce confidence limits on model simulations.

Physics-based modelling
An alternative approach to hydrological modelling is to seek to develop "physics-based
models," i.e. models explicitly based on the best available understanding of the physics of
hydrological processes. Such models are based on a continuum representation of catchment
processes and the equations of motion of the constituent processes are solved numerically
using a grid, of course discretized relatively crudely in catchment-scale applications. They
first became feasible in the 1970s when computing power became sufficient to solve the
relevant coupled Partial Differential Equations (Freeze and Harlan, 1969, Freeze, 1972). The
models are thus characterised by parameters that are in principle measurable and have a direct
physical significance. An important theoretical advantage is that if the physical parameters
can be determined a priori, such models can be applied to ungauged catchments, and the
effects of catchment change can be explicitly represented. However, whether this theoretical
advantage is achievable in practice is an open question at present.
 6

One of the best known models is the Systeme Hydrologique Europeen (SHE) model (Abbott
et al., 1986a,b), originally developed as a multinational European research collaboration. In
the UK this has been the subject of progressive development by the University of Newcastle
upon Tyne, and is known as the SHETRAN model (now including TRANsport of solutes and
sediments). A recent description is reported by Ewen et al. (2000). The catchment is
discretised on a grid square basis for the representation of land surface and subsurface
processes, creating a column of finite difference cells, which interact with cells from adjacent
columns to represent lateral flow and transport. River networks are modelled as networks of
stream links, with flow again represented by finite difference solution of the governing
equations. The resulting model is complex, computationally demanding and data intensive.
Ewen et al. (2000) note that a 1-year simulation typically has a 2 hour run time on an
advanced UNIX system.

In practice two fundamental problems arise with such models. The underlying physics has
been (necessarily) derived from small-scale, mainly laboratory-based, process observations.
Hence, firstly, the processes may not apply under field conditions and at field scales of
interest. There is, for example, numerical evidence that the effects of small-scale
heterogeneity may not be captured by effective, spatially-aggregated, properties (Binley and
Beven, 1989). Secondly, although the parameters may be measurable at small scale, they may
not be measurable at the scales of interest for application. An obvious example of both is the
representation of soil water flow at hillslope scale. Field soils are characterised by great
heterogeneity and complexity. Macropore flow is ubiquitous, yet neglected in physics-based
models, for lack of relevant theory and supporting data; the Richards' equation commonly
used for unsaturated flow depends on strongly non-linear functional relationships to represent
physical properties, for which there is no measurement basis at the spatial scales of practical
modelling interest. And field studies such as those of Pilgrim et al. (1978) demonstrate that
the dominant modes of process response cannot be specified a priori. For more detailed
discussion see, for example, Beven (1989).

There is, therefore, a need for fundamental research to address issues such as the appropriate
process representation and parameterisation at a given scale. For groundwater flow and
transport, significant progress has been made; new theoretical approaches to the
representation of heterogeneity have been developed (Dagan, 1986), and stochastic numerical
methods have been developed to represent explicitly the uncertainty associated with
heterogeneous properties (e.g. Wheater et al., 2000) and to incorporate conditioning on field
observations. Extension to the more complex problems of field-scale hydrology is urgently
needed, but severely constrained by data availability.

Most of the complexity of physically-based models, and the associated problems discussed
above, arise from the representation of subsurface flows, and the inherent lack of
observability of subsurface properties. The situation often met in arid areas is that overland
flow is the dominant runoff mechanism, and surface properties are in principle much more
readily obtained. It was therefore argued by Woolhiser thirty years ago (Woolhiser, 1971),
that it is in this environment that physics-based models are most likely to be successful. The
well-known KINEROS model is an outstanding example, and will be presented in its latest
form in the workshop paper by Darius Semmens et al.

HYDROLOGICAL PROCESSES IN ARID AREAS

 7

Despite the critical importance of water in arid and semi-arid areas, hydrological data have
historically been severely limited. It has been widely stated that the major limitation of the
development of arid zone hydrology is the lack of high quality observations (McMahon,
1979; Nemec and Rodier, 1979; Pilgrim et al., 1988). There are many good reasons for this.
Populations are usually sparse and economic resources limited; in addition the climate is
harsh and hydrological events infrequent, but damaging. However, in the general absence of
reliable long-term data and experimental research, there has been a tendency to rely on humid
zone experience and modelling tools, and data from other regions. At best, such results will
be highly inaccurate. At worst, there is a real danger of adopting inappropriate management
solutions which ignore the specific features of dryland response.

Despite the general data limitations, there has been some substantial and significant progress
in development of national data networks and experimental research. This has given new
insights and we can now see with greater clarity the unique features of arid zone hydrological
systems and the nature of the dominant hydrological processes. This provides an important
opportunity to develop methodologies for flood and water resource management which are
appropriate to the specific hydrological characteristics of arid areas and the associated
management needs, and hence to define priorities for research and hydrological data. The aim
here is to review this progress and the resulting insights, and to consider some of the
implications.

RAINFALL

Rainfall is the primary hydrological input, but rainfall in arid and semi-arid areas is
commonly characterised by extremely high spatial and temporal variability. The temporal
variability of point rainfall is well-known. Although most records are of relatively short
length, a few are available from the 19th century. For example, Table I presents illustrative
data from Muscat (Sultanate of Oman) (Wheater and Bell, 1983), which shows that a wet
month is one with one or two raindays. Annual variability is marked and observed daily
maxima can exceed annual rainfall totals.

For spatial characteristics, information is much more limited. Until recently, the major
source of detailed data has been from the South West U.S.A., most notably the two, relatively
small, densely instrumented basins of Walnut Gulch, Arizona (150km2) and Alamogordo
Creek, New Mexico (174km2), established in the 1950s (Osborn et al., 1979). The dominant
rainfall for these basins is convective; at Walnut Gulch 70% of annual rainfall occurs from
purely convective cells, or from convective cells developing along weak, fast-moving cold
fronts, and falls in the period July to September (Osborn and Reynolds, 1963). Raingauge
densities were increased at Walnut Gulch to give improved definition of detailed storm
structure and are currently better than 1 per 2km2. This has shown highly localised rainfall
occurrence, with spatial correlations of storm rainfall of the order of 0.8 at 2km separation,
but close to zero at 15-20km spacing. Osborn et al. (1972) estimated that to observe a
correlation of r2 = 0.9, raingauge spacings of 300-500m would be required.

Recent work has considered some of the implications of the Walnut Gulch data for
hydrological modelling. Michaud and Sorooshian (1994) evaluated problems of spatial
averaging for rainfall-runoff modelling in the context of flood prediction. Spatial averaging
on a 4kmx4km pixel basis (consistent with typical weather radar resolution) gave an
underestimation of intensity and led to a reduction in simulated runoff of on average 50% of
 8

observed peak flows. A sparse network of raingauges (1 per 20km2), representing a typical
density of flash flood warning system, gave errors in simulated peak runoff of 58%.
Evidently there are major implications for hydrological practice, and we will return to this
issue, below.

The extent to which this extreme spatial variability is characteristic of other arid areas has
been uncertain. Anecdotal evidence from the Middle East underlay comments that spatial and
temporal variability was extreme (FAO, 1981), but data from South West Saudi Arabia
obtained as part of a five-year intensive study of five basins (Saudi Arabian Dames and
Moore, 1988), undertaken on behalf of the Ministry of Agriculture and Water, Riyadh, have
provided a quantitative basis for assessment. The five study basins range in area from 456 to
4930 km2 and are located along the Asir escarpment (Fig 1), three draining to the Red Sea,
two to the interior, towards the Rub al Khali. The mountains have elevations of up to 3000m
a.s.l., hence the basins encompass a wide range of altitude, which is matched by a marked
gradient in annual rainfall, from 30-100mm on the Red Sea coastal plain to up to 450mm at
elevations in excess of 2000m a.s.l.

The spatial rainfall distributions are described by Wheater et al.(1991a). The extreme
spottiness of the rainfall is illustrated for the 2869km2 Wadi Yiba by the frequency
distributions of the number of gauges at which rainfall was observed given the occurrence of
a catchment rainday (Table 2). Typical inter-gauge spacings were 8-10km, and on 51% of
raindays only one or two raingauges out of 20 experienced rainfall. For the more widespread
events, sub-daily rainfall showed an even more spotty picture than the daily distribution. An
analysis of relative probabilities of rainfall occurrence, defined as the probability of rainfall
occurrence for a given hour at Station B given rainfall at Station A, gave a mean value of
0.12 for Wadi Yiba, with only 5% of values greater that 0.3. The frequency distribution of
rainstorm durations shows a typical occurrence of one or two-hour duration point rainfalls,
and these tend to occur in mid-late afternoon. Thus rainfall will occur at a few gauges and
die out, to be succeeded by rainfall in other locations. This is illustrated for Wadi Lith in
Figure 2, which shows the daily rainfall totals for the storm of 16th May 1984 (Fig2a), and
the individual hourly depths (Figs 2b-e). In general, the storm patterns appear to be consistent
with the results from the South West USA and area reduction factors were also generally
consistent with results from that region (Wheater et al., 1989).

The effects of elevation were investigated, but no clear relationship could be identified for
intensity or duration. However, a strong relationship was noted between the frequency of
raindays and elevation. It was thus inferred that once rainfall occurred, its point properties
were similar over the catchment, but occurrence was more likely at the higher elevations. It
is interesting to note that a similar result has emerged from a recent analysis of rainfall in
Yemen (UNDP, 1992), in which it was concluded that daily rainfalls observed at any location
are effectively samples from a population that is independent of position or altitude.

It is dangerous to generalise from samples of limited record length, but it is clear that most
events observed by those networks are characterized by extremely spotty rainfall, so much so
that in the Saudi Arabian basins there were examples of wadi flows generated from zero
observed rainfall. However, there were also some indications of a small population of more
wide-spread rainfalls, which would obviously be of considerable importance in terms of
surface flows and recharge. This reinforces the need for long-term monitoring of
experimental networks to characterise spatial variability.
 9

For some other arid or semi-arid areas, rainfall patterns may be very different. For example,
data from arid New South Wales, Australia have indicated spatially extensive, low intensity
rainfalls (Cordery et al., 1983), and recent research in the Sahelian zone of Africa has also
indicated a predominance of widespread rainfall. This was motivated by concern to develop
improved understanding of land-surface processes for climate studies and modelling, which
led to a detailed (but relatively short-term) international experimental programme, the
HAPEX-Sahel project based on Niamey, Niger (Goutorbe et al., 1997). Although designed to
study land surface/atmosphere interactions, rather than as an integrated hydrological study, it
has given important information. For example, Lebel et al. (1997) and Lebel and Le Barbe
(1997) note that a 100 raingauge network was installed and report information on the
classification of storm types, spatial and temporal variability of seasonal and event rainfall,
and storm movement. 80% of total seasonal rainfall was found to fall as widespread events
which covered at least 70% of the network. The number of gauges allowed the authors to
analyse the uncertainty of estimated areal rainfall as a function of gauge spacing and rainfall
depth.

Recent work in southern Africa (Andersen et al., 1998, Mocke, 1998) has been concerned
with rainfall inputs to hydrological models to investigate the resource potential of the sand
rivers of N.E.Botswana. Here, annual rainfall is of the order of 600mm, and available rainfall
data is spatially sparse, and apparently highly variable, but of poor data quality. Investigation
of the representation of spatial rainfall for distributed water resource modelling showed that
use of convential methods of spatial weighting of raingauge data, such as Theissen polygons,
could give large errors. Large sub-areas had rainfall defined by a single, possibly inaccurate
gauge. A more robust representation resulted from assuming catchment-average rainfall to
fall uniformly, but the resulting accuracy of simulation was still poor.

RAINFALL-RUNOFF PROCESSES

The lack of vegetation cover in arid and semi-arid areas removes protection of the soil from
raindrop impact, and soil crusting has been shown to lead to a large reduction in infiltration
capacity for bare soil conditions (Morin and Benyamini, 1977). Hence infiltration of
catchment soils can be limited. In combination with the high intensity, short duration
convective rainfall discussed above, extensive overland flow can be generated. This overland
flow, concentrated by the topography, converges on the wadi channel network, with the result
that a flood flow is generated. However, the runoff generation process due to convective
rainfall is likely to be highly localised in space, reflecting the spottiness of the spatial rainfall
fields, and to occur on only part of a catchment, as illustrated above.

Linkage between inter-annual variability of rainfall, vegetation growth and runoff production
may occur. Our modelling in Botswana suggests that runoff production is lower in a year
which follows a wet year, due to enhanced vegetation cover, which supports observations
reported by Hughes (1995).

Commonly, flood flows move down the channel network as a flood wave, moving over a bed
that is either initially dry or has a small initial flow. Hydrographs are typically characterised
by extremely rapid rise times, of as little as 15-30 minutes (Fig 3). However, losses from the
flood hydrograph through bed infiltration are an important factor in reducing the flood
volume as the flood moves downstream. These transmission losses dissipate the flood, and
obscure the interpretation of observed hydrographs. It is not uncommon for no flood to be
 10

observed at a gauging station, when further upstream a flood has been generated and lost to
bed infiltration.

As noted above, the spotty spatial rainfall patterns observed in Arizona and Saudi Arabia are
extremely difficult, if not impossible, to quantify using conventional densities of raingauge
network. This, taken in conjunction with the flood transmission losses, means that
conventional analysis of rainfall-runoff relationships is problematic, to say the least. Wheater
and Brown (1989) present an analysis of Wadi Ghat, a 597 km2 sub-catchment of wadi Yiba,
one of the Saudi Arabian basins discussed above. Areal rainfall was estimated from 5
raingauges and a classical unit hydrograph analysis was undertaken. A striking illustration of
the ambiguity in observed relationships is the relationship between observed rainfall depth
and runoff volume (Figure 4). Runoff coefficients ranged from 5.9 to 79.8%, and the greatest
runoff volume was apparently generated by the smallest observed rainfall! Goodrich et al.
(1997) show that the combined effects of limited storm areal coverage and transmission loss
give important differences from more humid regions. Whereas generally basins in more
humid climates show increasing linearity with increasing scale, the response of Walnut Gulch
becomes more non-linear with increasing scale. It is argued that this will give significant
errors in application of rainfall depth-area-frequency relationships beyond the typical area of
storm coverage, and that channel routing and transmission loss must be explicitly represented
in watershed modelling.

The transmission losses from the surface water system are a major source of potential
groundwater recharge. The characteristics of the resulting groundwater resource will depend
on the underlying geology, but bed infiltration may generate shallow water tables, within a
few metres of the surface, which can sustain supplies to nomadic people for a few months (as
in the Hesse of the North of South Yemen), or recharge substantial alluvial aquifers with
potential for continuous supply of major towns (as in Northern Oman and S.W. Saudi
Arabia).

The balance between localised recharge from bed infiltration and diffuse recharge from
rainfall infiltration of catchment soils will vary greatly depending on local circumstances.
However, soil moisture data from Saudi Arabia (Macmillan, 1987) and Arizona (Liu et al.,
1995), for example, show that most of the rainfall falling on soils in arid areas is subsequently
lost by evaporation. Methods such as the chloride profile method (e.g. Bromley et al., 1997)
and isotopic analyses (Allison and Hughes, 1978) have been used to quantify the residual
percolation to groundwater in arid and semi-arid areas.

In some circumstances runoff occurs within an internal drainage basin, and fine deposits can
support widespread surface ponding. A well known large-scale example is the Azraq oasis in
N.E. Jordan, but small-scale features (Qaa’s) are widespread in that area. Small scale
examples were found in the HAPEX-Sahel study (Desconnets et al., 1997). Infiltration from
these areas is in general not well understood, but may be extremely important for aquifer
recharge. Desconnets et al. report aquifer recharge of between 5 and 20% of basin
precipitation for valley bottom pools, depending on the distribution of annual rainfall.

The characteristics of the channel bed infiltration process are discussed in the following
section. However, it is clear that the surface hydrology generating this recharge is complex
and extremely difficult to quantify using conventional methods of analysis.

WADI BED TRANSMISSION LOSSES

 11

Wadi bed infiltration has an important effect on flood propagation, but also provides recharge
to alluvial aquifers. The balance between distributed infiltration from rainfall and wadi bed
infiltration is obviously dependant on local conditions, but soil moisture observations from
S.W. Saudi Arabia imply that, at least for frequent events, distributed infiltration of
catchment soils is limited, and that increased near surface soil moisture levels are
subsequently depleted by evaporation. Hence wadi bed infiltration may be the dominant
process of groundwater recharge. As noted above, depending on the local hydrogeology,
alluvial groundwater may be a readily accessible water resource. Quantification of
transmission loss is thus important, but raises a number of difficulties.

One method of determining the hydraulic properties of the wadi alluvium is to undertake
infiltration tests. Infiltrometer experiments give an indication of the saturated hydraulic
conductivity of the surface. However, if an infiltration experiment is combined with
measurement of the vertical distribution of moisture content, for example using a neutron
probe, inverse solution of a numerical model of unsaturated flow can be used to identify the
unsaturated hydraulic conductivity relationships and moisture characteristic curves. This is
illustrated for the Saudi Arabian Five Basins Study by Parissopoulos and Wheater (1992a).

In practice, spatial heterogeneity will introduce major difficulties to the up-scaling of point
profile measurements. The presence of silt lenses within the alluvium was shown to have
important effects on surface infiltration as well as sub-surface redistribution (Parissopoulos
and Wheater, 1990), and sub-surface heterogeneity is difficult and expensive to characterise.
In a series of two-dimensional numerical experiments it was shown that “infiltration
opportunity time”, i.e. the duration and spatial extent of surface wetting, was more important
than high flow stage in influencing infiltration, that significant reductions in infiltration occur
once hydraulic connection is made with a water table, and that hysteresis effects were
generally small (Parissopoulos and Wheater, 1992b). Also sands and gravels appeared
effective in restricting evaporation losses from groundwater (Parissopoulos and Wheater,
1991).

Additional process complexity arises, however. General experience from the Five Basins
Study was that wadi alluvium was highly transmissive, yet observed flood propagation
indicated significantly lower losses than could be inferred from in situ hydraulic properties,
even allowing for sub-surface heterogeneity. Possible causes are air entrapment, which could
restrict infiltration rates, and the unknown effects of bed mobilisation and possible pore
blockage by the heavy sediment loads transmitted under flood flow conditions.

A commonly observed effect is that in the recession phase of the flow, deposition of a thin (1-
2mm) skin of fine sediment on the wadi bed occurs, which is sufficient to sustain flow over
an unsaturated and transmissive wadi bed. Once the flow has ceased, this skin dries and
breaks up so that the underlying alluvium is exposed for subsequent flow events. Crerar et al.,
(1988) observed from laboratory experiments that a thin continous silt layer was formed at
low velocities. At higher velocities no such layer occurred, as the bed surface was mobilised,
but infiltration to the bed was still apparently inhibited. It was suggested that this could be
due to clogging of the top layer of sand due to silt in the infiltrating water, or formation of a
silt layer below the mobile upper part of the bed.

Further evidence for the heterogeneity of observed response comes from the observations of
Hughes and Sami (1992) from a 39.6 km2 semi-arid catchment in S.Africa. Soil moisture was
 12

monitored by neutron probe following two flow events. At some locations immediate
response (monitored 1day later) occurred throughout the profile, at others, an immediate
response near surface was followed by a delayed response at depth. Away from the inundated
area, delayed response, assumed due to lateral subsurface transmission, occurred after 21
days.

The overall implication of the above observations is that it is not possible at present to
extrapolate from in-situ point profile hydraulic properties to infer transmission losses from
wadi channels. However, analysis of observed flood flows at different locations can allow
quantification of losses, and studies by Walters (1990) and Jordan (1977), for example,
provide evidence that the rate of loss is linearly related to the volume of surface discharge.

For S.W. Saudi Arabia, the following relationships were defined:-

LOSSL = 4.56 + 0.02216 UPSQ - 2034 SLOPE + 7.34 ANTEC

(s.e. 4.15)
LOSSL = 3.75 x 10-5 UPSQ0.821 SLOPE -0.865 ACWW0.497
(s.e. 0.146 log units (±34%))

LOSSL = 5.7 x 10-5 UPSQ0.968 SLOPE -1.049

(s.e. 0.184 loge units (±44%))

Where:-

LOSSL = Transmission loss rate (1000m3/km) (O.R.1.08-87.9)

UPSQ = Upstream hydrograph volume (1000m3) (O.R. 69-3744)
SLOPE = Slope of reach (m/m) (O.R. 0.001-0.011)
ANTEC = Antecedent moisture index (O.R. 0.10-1.00)
ACWW = Active channel width (m) (O.R. 25-231)
and O.R. = Observed range

However, generalisation from limited experience can be misleading. Wheater et al. (1997)
analysed transmission losses between 2 pairs of flow gauges on the Walnut Gulch catchment
for a ten year sequence and found that the simple linear model of transmission loss as
proportional to upstream flow was inadequate. Considering the relationship:

Vx = V0 (1- α)x
 13

where Vx is flow volume (m3) at distance x downstream of flow volume V0 and α represents
the proportion of flow lost per unit distance, then α was found to decrease with discharge
volume:

α = 118.8 (V0)-0.71

The events examined had a maximum value of average transmission loss of 4076 m3 km-1 in
comparison with the estimate of Lane et al. (1971) of 4800-6700 m3 km-1 as an upper limit of
available alluvium storage.

The role of available storage was also discussed by Telvari et al. (1998), with reference to the
Fowler’s Gap catchment in Australia. Runoff plots were used to estimate runoff production as
overland flow for a 4km2 basin. It was inferred that 7000 m3 of overland flow becomes
transmission loss and that once this alluvial storage is satisfied, approximately two-thirds of
overland flow is transmitted downstream.

A similar concept was developed by Andersen et al. (1998) at larger scale for the sand rivers
of Botswana, which have alluvial beds of 20-200m width and 2-20m depth. Detailed
observations of water table response showed that a single major event after a seven weeks dry
period was sufficient to fully satisfy available alluvial storage (the river bed reached full
saturation within 10 hours). No significant drawdown occurred between subsequent events
and significant resource potential remained throughout the dry season. It was suggested that
two sources of transmission loss could be occurring, direct losses to the bed, limited by
available storage, and losses through the banks during flood events.

It can be concluded that transmission loss is complex, that where deep unsaturated alluvial
deposits exist the simple linear model as developed by Jordan (1977) and implicit in the
results of Walters (1990) may be applicable, but that where alluvial storage is limited, this
must be taken into account.

GROUNDWATER RECHARGE FROM EPHEMERAL FLOWS

The relationship between wadi flow transmission losses and groundwater recharge will
depend on the underlying geology. The effect of lenses of reduced permeability on the
infiltration process has been discussed and illustrated above, but once infiltration has taken
place, the alluvium underlying the wadi bed is effective in minimising evaporation loss
through capillary rise (the coarse structure of alluvial deposits minimises capillary effects).
Thus Hellwig (1973), for example, found that dropping the water table below 60cm in sand
with a mean diameter of 0.53mm effectively prevented evaporation losses, and Sorey and
Matlock (1969) reported that measured evaporation rates from streambed sand were lower
than those reported for irrigated soils.

Parrisopoulos and Wheater (1991) combined two-dimensional simulation of unsaturated

wadi-bed response with Deardorff’s (1977) empirical model of bare soil evaporation to show
that evaporation losses were not in general significant for the water balance or water table
response in short-term simulation (i.e. for periods up to 10 days). However, the influence of
vapour diffusion was not explicitly represented, and long term losses are not well understood.
Andersen et al. (1998) show that losses are high when the alluvial aquifer is fully saturated,
but are small once the water table drops below the surface.
 14

Sorman and Abdulrazzak (1993) provide an analysis of groundwater rise due to transmission
loss for an experimental reach in Wadi Tabalah, S.W. Saudi Arabia and estimate that on
average 75% of bed infiltration reaches the water table. There is in general little information
available to relate flood transmission loss to groundwater recharge, however. The differences
between the two are expected to be small, but will depend on residual moisture stored in the
unsaturated zone and its subsequent drying characteristics. But if water tables approach the
surface, relatively large evaporation losses may occur.

Again, it is tempting to draw over-general conclusions from limited data. In the study of the
sand-rivers of Botswana, referred to above, it was expected that recharge of the alluvial river
beds would involve complex unsaturated zone response. In fact, observations showed that the
first flood of the wet season was sufficient to fully recharge the alluvial river bed aquifer.
This storage was topped up in subsequent floods, and depleted by evaporation when the water
table was near-surface, but in many sections sufficient water remained throughout the dry
season to provide adequate sustainable water supplies for rural villages. And as noted above,
Wheater et al. (1997) showed for Walnut Gulch and Telvari et al. (1998) for Fraser’s Gap that
limited river bed storage affected transmission loss. It is evident that surface
water/groundwater interactions depend strongly on the local characteristics of the underlying
alluvium and the extent of their connection to, or isolation from, other aquifer systems.

Very recent work at Walnut Gulch (Goodrich et al., 2004) has investigated ephemeral
channel recharge using a range of experimental methods, combined with modelling. These
included a reach water balance method, including estimates of near channel
evapotranspiration losses, geochemical methods, analysis of changes in groundwater levels
and microgravity measurements, and unsaturated zone flow and temperature analyses. The
conclusions were that ephemeral channel losses were significant as an input to the underlying
regional aquifer, and that the range of methods for recharge estimation agreed within a factor
of three (reach water balance methods giving the higher estimates).

An important requirement for recharge estimation has arisen in connection with the proposal
for a repository for high level nuclear waste at Yucca Mountain, Nevada. Flint et al. (2002)
review a wide range of methods, including analysis of physical data from unsaturated zone
profiles of moisture and heat, environmental tracers, and watershed modelling. The results
indicate extreme variability in space and time, with watershed modelling giving a range from
zero to several hundred mm/year, depending on spatial location. The high values arise due to
flow focussing in ephemeral channels, and subsequent channel bed infiltration.

HYDROLOGICAL MODELLING AND THE REPRESENTATION OF RAINFALL

The preceding discussion illustrates some of the particular characteristics of arid areas which
place special requirements on hydrological modelling, for example for flood management or
water resources evaluation. One evident area of difficulty is rainfall, especially where
convective storms are an important influence. The work of Michaud and Sorooshian (1994)
demonstrated the sensitivity of flood peak simulation to the spatial resolution of rainfall
input. This obviously has disturbing implications for flood modelling, particularly where data
availability is limited to conventional raingauge densities. Indeed, it appears highly unlikely
that suitable raingauge densities will ever be practicable for routine monitoring. However, the
availability of 2km resolution radar data in the USA can provide adequate information and
radar could be installed elsewhere for particular applications. Morin et al. (1995) report
results from a radar located at Ben-Gurion airport in Israel, for example.
 15

One way forward is to develop an understanding of the properties of spatial rainfall based on
high density experimental networks and/or radar data, and represent those properties within a
spatial rainfall model for more general application. It is likely that this would have to be done
within a stochastic modelling framework in which equally-likely realisations of spatial
rainfall are produced, possibly conditioned by sparse observations.

Some simple empirical first steps in this direction were taken by Wheater et al. (1991a,b) for
S.W.Saudi Arabia and Wheater et al. (1995) for Oman. In the Saudi Arabian studies, as noted
earlier, raingauge data was available at approximately 10km spacing and spatial correlation
was low. Hence a multi-variate model was developed, assuming independence of raingauge
rainfall. Based on observed distributions, seasonally-dependent catchment rainday occurrence
was simulated, dependent on whether the preceding day was wet or dry. The number of
gauges experiencing rainfall was then sampled, and the locations selected based on observed
occurrences (this allowed for increased frequency of raindays with increased elevation).
Finally, start-times, durations and hourly intensities were generated. Model performance was
compared with observations. Rainfall from random selections of raingauges was well
reproduced, but when clusters of adjacent gauges were evaluated, a degree of spatial
organisation of occurrence was observed, but not simulated. It was evident that a weak degree
of correlation was present, which should not be neglected. Hence in extension of this
approach to Oman (Wheater et al., 1995), observed spatial distributions were sampled, with
satisfactory results.

However, this multi-variate approach suffers from limitations of raingauge density, and in
general a model in continuous space (and continuous time) is desirable. A family of
stochastic rainfall models of point rainfall was proposed by Rodriguez-Iturbe, Cox and Isham
(1987, 1988) and applied to UK rainfall by Onof and Wheater (1993,1994). The basic
concept is that a Poisson process is used to generate the arrival of storms. Associated with a
storm is the arrival of raincells, of uniform intensity for a given duration (sampled from
specified distributions). The overlapping of these rectangular pulse cells generates the storm
intensity profile in time. These models were shown to have generally good performance for
the UK in reproducing rainfall properties at different time-scales (from hourly upwards), and
extreme values.

Cox and Isham (1988) extended this concept to a model in space and time, whereby the
raincells are circular and arrive in space within a storm region. As before, the overlapping of
cells produces a complex rainfall intensity profile, now in space as well as time. This model
has been developed further by Northrop (1998) to include elliptical cells and storms and is
being applied to UK rainfall (Northrop et al., 1999).

Recent work (Samuel, 1999) has been exploring the capability of these models to reproduce
the convective rainfall of Walnut Gulch. In modelling point rainfall, the Bartlett-Lewis
Rectangular Pulse Model was generally slightly superior to other model variants tested. Table
3 shows representative performance of the model in comparing the hourly statistics from 500
realisations of July rainfall in comparison with 35 years from one of the Walnut Gulch
gauges (gauge 44).

where Mean is the mean hourly rainfall (mm), Var its variance, ACF1,2,3 the
autocorrelations for lags 1,2,3, Pwet the proportion of wet intervals, Mint the mean storm
 16

inter-arrival time (h), Mno the mean number of storms per month, Mdur the mean storm
duration (h).

This performance is generally encouraging (although the mean storm duration is

underestimated), and extreme value performance is excellent.

Work with the spatial-temporal model is still at a preliminary stage, but Fig 5 shows a
comparison of observed spatial coverage of rainfall for 25 years of July data from 81 gauges
(for different values of the standard deviation of cell radius) and Fig 6 the corresponding fit
for temporal lag-0 spatial correlation. Again, the results are encouraging, and there is promise
with this approach to address the significant problems of spatial representation for
hydrological modelling.

INTEGRATED MODELLING FOR WATER RESOURCE EVALUATION

Appropriate strategies for water resource development must recognise the essential physical
characteristics of the hydrological processes. Surface water storage, although subject to high
evaporation losses, is widely used, although temporal variability of flows must be adequately
represented to define long term yields. It can be noted that in some regions, for example, the
northern areas of southern Yemen, small scale storage has been developed as an appropriate
method to maximise the available resource from spatially-localised rainfall. Numbers of
small storages have been developed, some of which fill from localised rainfall. These then
provide a short-term resource for a nomadic family and its livestock.

Groundwater is a resource particularly well suited to arid regions. Subsurface storage

minimises evaporation loss and can provide long-term yields from infrequent recharge
events. The recharge of alluvial groundwater systems by ephemeral flows can provide an
appropriate resource, and this has been widely recognised by traditional development, such as
the “afalaj” of Oman and elsewhere. There may, however, be opportunities for augmenting
recharge and more effectively managing these groundwater systems. In any case, it is
essential to quantify the sustainable yield of such systems, for appropriate resource
development.

It has been seen that observations of surface flow do not define the available resource, and
similarly observed groundwater response does not necessarily indicate upstream recharge.
Figure 7 presents a series of groundwater responses from 1985/86 for Wadi Tabalah which
shows a downstream sequence of wells 3-B-96, -97, -98, -99 and -100 and associated surface
water discharges. It can be seen that there is little evidence of the upstream recharge at the
downstream monitoring point.

In addition, records of surface flows and groundwater levels, coupled with ill-defined
histories of abstraction, are generally insufficient to define long term variability of the
available resource.

To capture the variability of rainfall and the effects of transmission loss on surface flows, a
distributed approach is necessary. If groundwater is to be included, integrated modelling of
surface water and groundwater is needed. Distributed surface water models include
KINEROS (Wheater and Bell, 1983, Michaud and Sorooshian, 1994) and the model of
Sharma (1997, 1998). A distributed approach to the integrated modelling of surface and
groundwater response following Wheater et al.(1995) is illustrated in Fig 8. This requires the
 17

characterisation of the spatial and temporal variability of rainfall, distributed infiltration,

runoff generation and flow transmission losses, the ensuing groundwater recharge and
groundwater response. This presents some technical difficulties, although the integration of
surface and groundwater modelling allows maximum use to be made of available
information, so that, for example, groundwater response can feed back information to
constrain surface hydrological parameterisation. It does, however, provide the only feasible
method of exploring the internal response of a catchment to management options.

In a recent application, this integrated modelling approach was developed for Wadi Ghulaji,
Sultanate of Oman, to evaluate options for groundwater recharge management (Wheater et
al., 1995). The catchment, of area 758 km2, drains the southern slopes of Jebal Hajar in the
Sharqiyah region of Northern Oman. Proposals to be evaluated included recharge dams to
attenuate surface flows and provide managed groundwater recharge in key locations. The
modelling framework involved the coupling of a distributed rainfall model, a distributed
water balance model (incorporating rainfall-runoff processes, soil infiltration and wadi flow
transmission losses), and a distributed groundwater model (Fig 9).

The representation of rainfall spatial variability presents technical difficulties, since data are
limited. Detailed analysis was undertaken of 19 rain gauges in the Sharqiyah region, and of
six raingauges in the catchment itself. A stochastic multi-variate temporal- spatial model was
devised for daily rainfall, a modified version of a scheme orgininally developed by Wheater
et al., 1991a,b. The occurrence of catchment rainfall was determined according to a
seasonally-variable first order markov process, conditioned on rainfall occurrence from the
previous day. The number and locations of active raingauges and the gauge depths were
derived by random sampling from observed distributions.

The distributed water balance model represents the catchment as a network of two-
dimensional plane and linear channel elements. Runoff and infiltration from the planes was
simulated using the SCS approach. Wadi flows incorporate a linear transmission loss
algorithm based on work by Jordan (1977) and Walters (1990). Distributed calibration
parameters are shown in Figure 10.

Finally, a groundwater model was developed based on a detailed hydrogeological

investigation which led to a multi-layer representation of uncemented gravels,
weakly/strongly cemented gravels and strongly cemented/fissured gravel/bedrock, using
MODFLOW.

The model was calibrated to the limited flow data available (a single event) (Table 4), and
was able to reproduce the distribution of runoff and groundwater recharge within the
catchment through a rational association on loss parameters with topography, geology and
wadi characteristics. Extended synthetic data sequences were then run to investigate
catchment water balances under scenarios of different runoff exceedance probabilities (20%,
50%, 80%), as in Table 4, and to investigate management options.

CONCLUSIONS

It has been shown that for many applications, the hydrological characteristics of arid areas
present severe problems for conventional methods of analysis. Recent data are providing
new insights. These insights must be used as the basis for development of more appropriate
methods for flood design and water resource evaluation, and in turn, to define data needs and
 18

research priorities. Much high quality research is needed, particularly to investigate

processes such as spatial rainfall, and infiltration and groundwater recharge from ephemeral
flows.

For developments to maximise the resource potential, define long-term sustainable yields and
protect traditional sources, it is argued that distributed modelling is a valuable, if not essential
tool. However, this confronts severe problems of characterisation of rainfall, rainfall-runoff
processes, and groundwater recharge, and of understanding the detailed hydrogeological
response of what are often complex groundwater systems. Similarly, new approaches to
flood design and management are required which represent the extreme value characteristics
of arid areas and recognise the severe problems of conventional rainfall-runoff analysis.

Above all, basic requirements are for high quality data of rainfall, surface water flows and
groundwater response to support regional analyses and the development of appropriate
methodologies. Too often, studies focus on either surface or subsurface response without
taking an integrated view. Too often, networks are reduced after a few years without
recognition that the essential variability of wadi response can only be characterised by
relatively long records. Quality control of data is vital, but can easily be lost sight of with
ready access to computerised data-bases.

Superimposed on these basic data needs are the requirement for specific process studies,
including sediment transport, surface water/groundwater interactions in the active wadi
channel, evaporation processes and consumptive use of wadi vegetation, and the wider issues
of groundwater recharge. These are challenging studies, with particularly challenging
logistical problems, and require the full range of advanced hydrological experimental
methods to be applied, particularly integrating quantity and quality data to deduce system
responses, and making full use of remote sensing and geophysical methods to characterise
system properties.

It must not be forgotten that in general, data networks are under threat world-wide, and a
major priority for hydrologists must be to promote recognition of the value of data for water
management, the importance of long records in a region characterised by high inter-annual
variability, and of the particular technical and logistical difficulties in capturing hydrological
response in arid areas. The current International Hydrological Programme rightly prioritizes
hydrological data as the essential foundation for effective management. The results of both
detailed research and regional analyses are required for the essential understanding of wadi
hydrology which must underlie effective management.

REFERENCES

Abbott, M.B., Bathurst, J. C., Cunge, J.A., O'Connell, P. E. and Rasmussen, J. (1986a) An
introduction to the European Hydrological System - Systeme Hydrologique Europeen, SHE.
1. History and philosophy of a physically-based, distributed modelling system. J. Hydrol., 87,
45-59
 19

Abbott, M.B., Bathurst, J. C., Cunge, J.A., O'Connell, P. E. and Rasmussen, J. (1986b)An
introduction to the European Hydrological System - Systeme Hydrologique Europeen, SHE.
2. Structure of a physically-based, distributed modelling system. J. Hydrol., 87, 61-77

Allison, G.B. and Hughes, M.W. (1978) The use of environmental chloride and tritium to
estimate total recharge to an unconfined aquifer. Aust.J.Soil Res.,16, 181-195

Andersen, N.J., Wheater, H.S., Timmis, A.J.H. and Gaongalelwe, D. (1998) Sustainable
development of alluvial groundwater in sand rivers of Botswana. In Sustainability of Water
Resources under Increasing Uncertainty, IAHS Pubn. No. 240, pp 367-376.

Beven, K.J. Changing ideas in hydrology: the case of physically-based models. J.Hydrol.,
105, 157-172 (1989)

Beven, K.J. (1993) Prophecy, reality and uncertainty in distributed hydrological modelling.
Adv. In Water Resourc., 16, 41-51

Beven, K.J. (2000) Rainfall-Runoff Modelling - the Primer, Wiley.

Beven, K.J. and Binley, A.M. (1992) The future of distributed models: model calibration and
predictive uncertainty. Hydrol. Processes, 6, 279-298

Binley, A.M. and Beven, K.J. (1989)A physically based model of heterogeneous hillslopes 2.
Effective hydraulic conductivities. Water Resour. Res.,25, 6, 1227-1233

Bromley, J., Edmunds, W.M., Fellman E., Brouwer, J., Gaze, S.R., Sudlow, J. and Taupin, J.-
D. (1997) Estimation of rainfall inputs and direct recharge to the deep unsaturated zone of
southern Niger using the chloride profile method. J. Hydrol. 188-189, pp.139-154

Cordery, I., Pilgrim, D.H and Doran, D.G. (1983) Some hydrological characteristics of arid
western New South Wales. The Institution of Engineers, Australia, Hydrology and Water
Resources Symp, Nov.

Cox, D. R. and Isham, V.(1988) A simple spatial-temporal model of rainfall, Proc.Roy.Soc.,

A415,317-328

Crawford, N.H. and Linsley, R.K. (1966) Digital simulation in hydrology: Stanford
Watershed Model IV. Tech. Rpt 39, Stanford University, California

Crerar, S., Fry, R.G., Slater, P.M., van Langenhove, G. and Wheeler, D. (1988) An
unexpected factor affecting recharge from ephemeral river flows in SWA/Namibia. In
Estimation of Natural Groundwater Recharge, I.Simmers (ed.), pp.11-28, D.Reidel
Publishing Company.

Dagan, G. (1986) Statistical theory of groundwater flow and transport: pore to laboratory;
laboratory to formation and formation to regional scale. Wat. Resour. Res., 22, 120-135

Deardorff, J.W. (1977) A parameterization of ground-surface moisture content for use in

atmospheric prediction models. J. Applied. Meteorol., 16, 1182-1185.
 20

Desconnets, J.C., Taupin, J.D., Lebel, T. and Leduc, C. (1997) Hydrology of the HAPEX-
Sahel Central Super-Site: surface water drainage and aquifer recharge through the pool
systems. J. Hydrol. 188-189, pp.155-178

Duan, Q., Gupta, H.V., Sorooshian, S., Rousseau, A.N., Turcotte, R. (eds.) (2003)
Calibration of Watershed Models. Water Science and Application Vol 6, Ed Qingyun Duan,
Gupta, H.V., Sorooshian, S., Rousseau, A.N., Turcotte, R. American Geophysical Union.

Eagleson, P.S., Fennessey, N.M., Qinliang, W. and Rodriguez-Iturbe, I., (1987) Application
of spatial Poisson models to air mass thunderstorm rainfall. J. Geophys. Res., 92: 9661-9678.

Ewen, J, Parkin, G, and O'Connell, P.E. SHETRAN: distributed river basin flow and
transport modeling system. Journal of Hydrologic Engineering, July, 250-258 (2000)

Flint, A.L., Flint, L.E., Kwicklis, E.M., Fabryka-Martin, J.T. and Bodvarsson, G.S. (2002)
Estimating recharge at Yucca Mountain, Nevada, USA: comparison of methods.
Hydrogeology Journal 10:180-204

Food and Agriculture Organization, (1981) Arid zone hydrology for agricultural
development. F.A.O, Rome, Irrig. Drain. Pap. 37, 271pp.

Freer, J., Beven, K. and Abroise, B. (1996) Bayesian uncertainty in runoff prediction and the
value of data: an application of the GLUE approach. Water Resour. Res. 32, 2163-2173

Freeze, R.A. (1972) Role of subsurface flow in generating surface runoff. 2:Upstream source
areas. Water Resour. Res.,8,5, 1272-1283

Freeze, R.A. and Harlan, R.L. (1969) Blueprint for a physically-based, digitally simulated
hydrologic response model. J. Hydrol, 9, 237-258

Goodrich, D.C., Lane, L.J., Shillito, R.M., Miller, S.N., Syed, K.H. and Woolhiser, D.A.
(1997) Linearity of basin response as a function of scale in a semi-arid watershed. Water
Resour. Res., 33,12, 2951-2965.

Goodrich, D.C., Williams, D.G., Unkrich, C.L., Hogan, J.F., Scott, R.L., Hultine, K.R., Pool,
D., Coes, A.L. and Miller, S. (2004) Comparison of methods to Estimate Ephemeral Channel
Recharge, Walnut Gulch, San Pedro River Basin, Arizona. In: Groundwater Recharge in a
Desert Environment. The Southwestern United States, Eds. Hogan, J.F., Phillips, F.M. and
Scanlon, B.R. Water Science and Application 9, American Geophysical Union

Goutorbe, J.P., Dolman, A.J., Gash, J.H.C., Kerr, Y.H., Lebel, T., Prince S.D., and Stricker,
J.N.M. (1997) (Eds). HAPEX-Sahel. Elsevier, 1079 pp.(reprinted from J.Hydrol. 188-189/1-
4)

Gupta, H.V., Sorooshian, S. and Yapo, P.O. (1998) Towards improved calibration of
hydrological models: multiple and non-commensurable measures of information. Water
Resour. Res. 34(4), 751-763

Hellwig, D.H.R. (1973) Evaporation of water from sand, 3: The loss of water into the
atmosphere from a sandy river bed under arid climatic condtions. J. Hydrol., 18, 305-316.
 21

Hughes, D.A. (1995) Monthly rainfall-runoff models applied to arid and semiarid catchments
for water resource estimation purposes. Hydr. Sci. J.,40, 6, 751-769.

Hughes, D.A. and Sami, K., (1992) Transmission losses to alluvium and associated moisture
dynamics in a semiarid ephemeral channel system in Southern Africa. Hydrological
Processes, 6, 45-53

Institute of Hydrology. Flood Estimation Handbook (1999) 5 vols, Wallingford UK.

Jacobs, B.L., Rodrigues-Iturbe, I. and Eagleson, P.S., (1988) Evaluation of a homogeneous

point process description of Arizona thunderstorm rainfall. Water Resour. Res., 24(7): 1174-
1186.

Jordan, P.R. (1977) Streamflow transmission losses in Western Kansas. Jul of Hydraulics
Division, ASCE, 108, HY8, 905-919.

Kleissen, F.M., Beck, M.B. & Wheater, H.S. (1990) "Identifiability of conceptual
hydrochemical models". Water Resources Research, 26, 2 2979-2992.

Lane, L.J., Diskin, M.H. and Renard, K.G. (1971) Input-output relationships for an ephemeral
stream channel system. J.Hydrol. 13, pp.22-40.

Lebel, T., Taupin, J.D., and D’Amato, N. (1997) Rainfall monitoring during HAPEX-
Sahel.1. General rainfall conditions and climatology. J.Hydrol. 188-189,pp.74-96

Lebel, T., and Le Barbe, L. (1997) Rainfall monitoring during HAPEX-Sahel.2. Point and
areal estimation at the event and seasonal scales. J.Hydrol. 188-189, pp.97-122

Lees, M.J. and Wagener, T. (1999) A monte-carlo analysis toolbox (MCAT) for Matlab -
User manual. Imperial College

Liu, B., Phillips, F., Hoines, S., Campbell, A.R., Sharma,P. (1995) Water movement in desert
soil traced by hydrogen and oxygen isotopes, chloride, and chlorine-36, southern Arizona.
J.Hydrol. 168, pp.91-110.

Macmillan, L.C. (1987) Regional evaporation and soil moisture analysis for irrigation
application in arid areas. Univ London MSc thesis, Dept of Civil Engineering, Imperial
College. 223pp

Mocke, R. (1998) Modelling the sand-rivers of Botswana – distributed modelling of runoff

and groundwater recharge processes to assess the sustainability of rural water supplies. Univ
London MSc thesis, Dept of Civil Engineering, Imperial College. 148pp

Michaud, J.D and Sorooshian, S (1994) Effect of rainfall-sampling errors on simulations of

desert flash floods. Water Resour.Res., 30, 10, pp. 2765-2775.

Morin, J. and Benyamini, Y. (1977) Rainfall infiltration into bare soils. Water. Resour. Res.,
13, 5, 813-817.
 22

Morin,J., Rosenfeld,D. and Amitai, E. (1995) Radar rain field evaluation and possible use of
its high temporal and spatial resolution for hydrological purposes. J.Hydrol., 172, 275-292.

McMahon, T.A., (1979) Hydrological characteristics of arid zones. Proceedings of a

Symposium on The Hydrology of Areas of Low Precipitation, Canberra, IAHS Publ. No.
128, pp. 105-123.

Natural Environment Research Council (1975) Flood Studies Report, 5 Vols, Wallingford,
UK

Nemec, J. and Rodier, J.A., (1979) Streamflow characteristics in areas of low precipitation.
Proceedings of a Symposium on the Hydrology of Areas of Low Precipitation, Canberra,
IAHS Publ. No. 128, pp. 125-140.

Northrop, P. J. (1998), A clustered spatial-temporal model of rainfall.

Proc.Roy.Soc.,A454,1875-1888

Northrop, P.J., Chandler, R.E., Isham, V.S., Onof, C. and Wheater, H.S. (1999) Spatial-
temporal stochastic rainfall modelling for hydrological design. In: Hydrological Extremes:
Understanding, Predicting, Mitigating. Eds. Gottschalk, L., Olivry, J.-C., Reed, D. and
Rosbjerg, D. IAHS Publn. No. 255, pp 225-235.

Onof, C. and Wheater, H.S. (1993) "Modelling of British rainfall using a random parameter
Bartlett-Lewis rectangular pulse model" J. Hydrol. 149, 67-95.

Onof, C. and Wheater, H.S. (1994) "Improvements of the modelling of British rainfall using
a modified random parameter Bartlett-Lewis rectangular pulse model" J. Hydrol 157 177-
195.

Osborn, H.B. and Reynolds, W.N. (1963) Convective Storm Patterns in the Southwestern
United States. Bull. IASH 8(3) 71-83

Osborn, H.B., Renard, K.G. and Simanton, J.R., (1979) Dense networks to measure
convective rainfalls in the Southwestern United States. Water Resour. Res.,15(6):1701-1711.

Osborn, H.B., Lane, L.J. and Myers, V.A. (1980) Rainfall/Watershed Relationships for
Southwestern Thunderstorms. Trans ASAE 23(1) 82-91

Parissopoulos, G.A. and Wheater, H.S. (1990) Numerical study of the effects of layers on
unsaturated-saturated two-dimensional flow. Water Resources Mgmt 4, 97-122.

Parissopoulos, G.A. & Wheater, H.S. (1991) Effects of evaporation on groundwater recharge
from ephemeral flows. In: Advances in Water Resources Technology, Ed. G. Tsakiris, A.A.
Balkema, 235-245, 1991.

Parrisopoulos, G.A. & Wheater, H.S. (1992a) Experimental and numerical infiltration studies
in a wadi stream-bed. J. Hydr. Sci. 37, 27-37.
 23

Parissopoulos, G.A. & Wheater H.S. (1992b) Effects of hysteresis on groundwater recharge
from ephemeral flows, Water Resour. Res., 28, 11, 3055-3061.

Pilgrim, D.H., Chapman, T.G. and Doran, D.G., (1988) Problems of rainfall-runoff
modelling in arid and semi-arid regions. Hydrol. Sci. J., 33(4): 379-400.

Pilgrim, D.H., Huff, D.D. and Steele, T.D. (1978) A field evaluation of subsurface and
surface runoff. J Hydrol., 38, 319-341

Reid, I., and Frostick, L.E. (1987) Flow dynamics and suspended sediment properties in arid
zone flash floods. Hydrol. Processes, 1,3, 239-253.

Rodriguez-Iturbe, I., Cox, D. R. and Isham, V. (1987) Some models for rainfall based on
stochastic point processes.Proc.Roy.Soc., A410, 269-288,

Rodriguez-Iturbe, I., Cox, D. R. and Isham, V. (1988) A point process model for rainfall:
further developments. Proc.Roy.Soc.,A417, 283-298

Samuel, C.R. (1999) Stochastic Rainfall Modelling of Convective Storms in Walnut Gulch,
Arizona. Univ London PhD Thesis, 235pp

Saudi Arabian Dames and Moore, (1988) Representative Basins Study. Final Report to
Ministry of Agriculture and Water, Riyadh, 84 vols.

Sharma, K.D. (1997) Integrated and sustainable development of water resources of the Luni
basin in the Indian arid zone. In Sustainability of Water Resources under Increasing
Uncertainty, IAHS Publn. No.240, pp 385-393.

Sharma, K.D. (1998) Resource assessment and holistic management of the Luni River Basin
in the Indian desert. In Hydrology in a changing environment, Vol II, Eds. Howard Wheater
and Celia Kirby, Wiley. pp 387-395

Singh, V.P. and Frevert, D.K. (eds.) (2002a) Mathematical models of large watershed
hydrology. Water Resources Publications, LLC.

Singh, V.P. and Frevert, D.K. (eds.) (2002b) Mathematical models of small watershed
hydrology and applications. Water Resources Publications, LLC.

Sorey, M.L. and Matlock, W.G. (1969) Evaporation from an ephemeral streambed. J.
Hydraul. Div. Am. Soc. Civ. Eng., 95, 423-438.

Sorman, A.U. and Abdulrazzak, M.J. (1993) Infiltration - recharge through wadi beds in arid
regions. Hydr. Sci. Jnl., 38, 3, 173-186.

Spear, R.C. and Hornberger, G.M. (1980) Eutrophication in Peel inlet, II, Identification of
critical uncertainties via generalised sensitivity analysis. Water Resour. Res. 14, 43-49
 24

Telvari, A., Cordery, I. and Pilgrim, D.H. (1998) Relations between transmission losses and
bed alluvium in an Australian arid zone stream. In Hydrology in a Changing Environment.
Eds Howard Wheater and Celia Kirby,Wiley, Vol II, pp 361-36

Travers Morgan (1993) Detailed hydrotechnical recharge studies in Wadi Ghulaji. Final
Report to Ministry of Water Resources, Oman.

UNDP (1992) Surface Water Resources. Final report to the Government of the Republic of
Yemen High Water Council UNDP/DESD PROJECT YEM/88/001 Vol III, June.

Wagener, T., Lees, M.J. and Wheater, H.S. (1999) A rainfall-runoff modelling toolbox
(RRMT) for Matlab - User manual. Imperial College

Wagener, T., Boyle, D.P., Lees, M.J., Wheater, H.S., Gupta, H.V. and Sorooshian, S.
A framework for the development and application of hydrological models. Proc BHS 7th
National Symp., Newcastle upon Tyne, pp 3.75-3.81 (2000)

Wagener, T., Lees, M.J. and Wheater, H.S. (2002) A toolkit for the development and
application of parsimonious hydrological models. In Singh, Frevert and Meyer (eds.)
Mathematical models of small watershed hydrology-Volume 2. Water Resources Publications
LLC, USA

Wagener, T., McIntyre, N., Lees, M.J., Wheater, H.S. and Gupta, H.V. (2003a) Towards
reduced uncertainty in conceptual rainfall-runof modelling: Dynamic Identifiability Analysis.
Hydrological Processes,17, 455-476

Wagener, T. Wheater, H.S. and Gupta, H.V. (2003b) Identification and evaluation of
conceptual rainfall-runoff models. In: Duan, q., Sorooshian, S., Gupta, H.V. Rousseau, A.
and Turcotte, R. (Eds.) Advances in calibration of watershed models. AGU MOnograpg
Series, USA, 29-47

Wagener, T. Wheater, H.S. and Gupta, H.V. (2004) Rainfall-Runoff Modelling in Gauged
and Ungauged Catchments. Imperial College Press, 306pp

Walters, M.O. (1990) Transmission losses in arid region. Jnl of Hydraulic Engineering, 116,
1, 127-138.

Wheater, H.S. and Bell, N.C. (1983) Northern Oman flood study. Proc. Instn. Civ. Engrs.
Part II, 75, 453-473.

Wheater, H.S. & Brown , R.P.C. (1989) Limitations of design hydrographs in arid areas - an
illustration from southwest Saudi Arabia. Proc. 2nd Natl. BHS Symp. (1989), 3.49-3.56.

Wheater, H.S., Bishop, K.H. & Beck, M.B. (1986) "The identification of conceptual
hydrological models for surface water acidification". J. Hydrol. Proc. 1 89-109.

Wheater, H.S., Larentis, P. and Hamilton, G.S., (1989) Design rainfall characteristics for
southwest Saudi Arabia. Proc. Inst. Civ. Eng., Part 2, 87: 517-538.
 25

Wheater, H.S., Butler, A.P., Stewart, E.J. & Hamilton, G.S. (1991a) A multivariate spatial-
temporal model of rainfall in S.W. Saudi Arabia. I. Data characteristics and model
formulation. J. Hydrol., 125, 175-199.

Wheater, H.S., Onof, C., Butler, A.P. and Hamilton, G.S., (1991b) A multivariate spatial-
temporal model of rainfall in southwest Saudi Arabia. II Regional analysis and long-term
performance. J. Hydrol., 125, 201-220.

Wheater, H.S., Jakeman, A.J. and Beven, K.J., (1993) Progress and directions in rainfall-
runoff modelling. In: Modelling Change in Environmental Systems, Ed. A.J. Jakeman, M.B.
Beck and M.J. McAleer, Wiley, 101-132.

Wheater, H.S., Jolley, T.J. and Peach, D. (1995) A water resources simulation model for
groundwater recharge studies: an application to Wadi Ghulaji, Sultanate of Oman. In: Proc.
Intnl. Conf. on Water Resources Management inArid Countries (Muscat), 502-510.

Wheater, H.S., Woods Ballard, B. and Jolley, T.J., (1997) An integrated model of arid zone
water resources: evaluation of rainfall-runoff simulation performance. In: Sustainability of
Water Resources under Increasing Uncertainty, IAHS Pubn. No. 240, pp 395-405.

Wheater, H.S., Tompkins, M.A., van Leeuwen, M. and Butler, A.P. (2000) Uncertainty in
groundwater flow and transport modelling - a stochastic analysis of well protection zones.
Hydrol. Proc., 14, 2019-2029.

Woolhiser, D.A. (1971) Deterministic approach to watershed modelling. Nordic Hydrology

11, 146-166
 26

Figure 1. Location of Saudi Arabian study basins

 27

Figure 2a. Wadi AL-Lith daily rainfall, 16th May 1994

 28

Figure 2b. Wadi Al-Lith hourly rainfall, 16th May 1994

 29

Figure 2c. Wadi Al-Lith hourly rainfall, 16th May 1994

 30

Figure 2d. Wadi Al-Lith hourly rainfall, 16th May 1994

 31

Figure 2e. Wadi Al-Lith hourly rainfall, 16th May 1994

 32

Figure 3. Surface water hydrographs, Wadi Ghat 12 May 1984 –

observations and unit hydrograph simulation
 33

Figure 4. Storm runoff as a function of rainfall, Wadi Ghat

0.9

0.8

0.7

0.6
Data
Coverage

1
0.5 2
5
10
0.4

0.3

0.2

0.1

0
0.75 0.8 0.85 0.9 0.95 1
Frequency
 34

Figure 5. Frequency distribution of spatial coverage of Walnut Gulch rainfall.

Observed vs. alternative simulations
1

0.9

0.8

0.7

0.6
Data
1
0.5 2
5
10
0.4

0.3

0.2

0.1

0
0 5000 10000 15000 20000 25000
Distance (m)

Figure 6. Spatial correlation of Walnut Gulch rainfall.

Observed vs. alternative simulations
 35

Figure 7. Longitudinal sequence of wadi alluvium well hydrographs and

associated surface flows, Wadi Tabalah, 1985/6

DATA MODELS

RAINFALL DATA STOCHASTIC SPATIAL

ANALYSIS RAINFALL SIMULATION

RAINFALL RUNOFF
CALIBRATION DISTRIBUTED RAINFALL-RUNOFF
MODEL

WABI BED TRANSMISSION LOSS

GROUNDWATER RECHARGE

GROUNDWATER
CALIBRATION DISTRIBUTED GROUNDWATER
MODEL

Figure 8. Integrated modelling strategy for water resource evaluation

 36

RAINFALL GENERATOR

JEBEL PLANE
Water Balance Model
WADI BED
ALLUVIAL PLANE
Water Balance Model Transmission Lose
Model

GROUNDWATER MODEL

Figure 9 Schematic of the distributed water resource model

Figure 10. Distributed calibration parameters, water balance model

 37

Table 1. Summary of Muscat rainfall data (1893 -1959) (after Wheater and Bell, 1983)

Monthly rainfall (mm) Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.

Mean 31.2 19.1 13.1 8.0 0.38 1.31 0.96 0.45 0.0 2.32 7.15 22.0

Standard deviation 38.9 25.1 18.9 20.3 1.42 8.28 4.93 2.09 0.0 7.62 15.1 35.1

Max. 143.0 98.6 70.4 98.3 8.89 64.0 37.1 14.7 0.0 44.5 77.2 171.2

Mean number of raindays 2.03 1.39 1.15 0.73 0.05 0.08 0.10 0.07 0.0 0.13 0.51 1.6

Max. daily fall (mm) 78.7 57.0 57.2 51.3 8.9 61.5 30.0 10.4 0.0 36.8 53.3 57.2

Number of years record 63.0 64 62 63 61 61 60 61 61 60 61 60

 38

Table 2. Wadi Yiba raingauge frequencies and associated conditional

probabilities for catchment rainday occurrence

Number of Gauges Occurrence Probability

1 88 0.372
2 33 0.141
3 25 0.106
4 18 0.076
5 10 0.042
6 11 0.046
7 13 0.055
8 6 0.026
9 7 0.030
10 5 0.021
11 7 0.030
12 5 0.021
13 3 0.013
14 1 0.004
15 1 0.004
16 1 0.005
17 1 0.004
18 1 0.004
19 0 0.0
20 0 0.0
TOTAL 235 1.000

Table 3. Performance of the Bartlett-Lewis Rectangular Pulse Model in representing July

rainfall at gauge 44, Walnut Gulch
Mean Var ACF1 ACF2 ACF3 Pwet Mint Mno Mdur

Model 0.103 1.082 0.193 0.048 0.026 0.032 51.17 14.34 1.68

Data 0.100 0.968 0.174 0.040 0.036 0.042 53.71 13.23 2.38
 39

Table 4. Annual catchment water balance, simulated scenarios

Groundwater
Scenario Rainfall Evaporation Runoff %Runoff
recharge
Wet 88 0.372 12.8 4.0 4.6

Average 33 0.141 11.2 3.5 4.0

Dry 25 0.106 5.5 1.7 3.2